But to figure out the resistance ratio, I need to know the resistance of my PIR sensor in reverse polarity situation. If the sensor normally draws 1uA at 5V, does that mean that it has 5 megaOhms of resistance? Do electronic components have fixed internal resistance?
No, electrical components do not have fixed internal resistance. Even if some current does flow, it is really unlikely to cause any harm if the power involved is in the nanowatt range.
Does the requirement to use shielded cables mean that I can't use the standard breadboard jumper wires?
So this is why I warned you that you might be worrying about the wrong things. First, one of those datasheets lists a maximum supply voltage of 4.5VDC, so how are you handling that?
I was thinking of using a Raspberry Pi Pico to provide 3.3V to the EKMBs.
The Pico has a RT6150B buck-boost converter with a current limit of 800mA and a power dissipation of 2-3W, which should be enough so that it doesn't overheat.
But now I think about it, the RT6150B might fail too? If it fails short, then it will pass the 5V USB power directly to the EKMB which can only tolerate 4.5V.
And what if the RT6150B fails in such a way that it drives up the voltage very high? Maybe I should put a shunt zener diode to prevent over-voltage here too? I guess I should also add a fuse.
I also could not find information on the RT6150B datasheet about its noise and voltage regulation.
If you are using the other one, it says 7.0VDC and that seems like the basic circuit above is perfectly adequate. However, note that it specifies a maximum of 200V static electricity and shielded cables. This is very, very low and indicates that these things are delicate and sensitive to transients. A breadboard or breadboard jumpers will give you pretty bad results. Wall-wart leakage may be an issue, as I mentioned earlier. You need soldered twisted pairs inside a shield or coax or something like that. And ground the negative supply rail properly.
I found this YouTUbe video:
In that video, it looks like the guy (Andreas Spiess) is just using a regular ESP32 with some regular jumper cables to attach to the PaPIRs. It looks like the jumper cables are attached via some heat-shrink wrap? Did he solder the pin to the exposed jumper wire first and then use heat-shrink tubing to wrap it?
Anyway, that video is what gave me the idea that it is probably okay to connect a PaPIR to directly to an Arduino Uno or a Pi Pico via some jumper cables, but then I got worried about the voltage regulator failing on the Arduino Uno and that's what made me ask about all this.
If using coaxial jumper wires, I would have to ground the shielding, right? Seems really complicated for a jumper wire that will be at most 20cm long.
Is grounding something just means connecting it to -ve? I'm thinking of connecting my power supply to the breadboard using one of those barrel jack screw terminal things - I'm not sure what the rating is on these things, can never find a datasheet, but most places say they're okay for 1-2A which should be plenty for my needs. Anyway, the screw terminal has a +ve and -ve coming out of it, so grounding something just means connecting it to the -ve, right?